REPORT 6 OF THE COUNCIL ON SCIENCE AND PUBLIC HEALTH (A-12) Screening Mammography (Resolution 509-A-10, Resolve 1) (Reference Committee E)

Transcription

1 REPORT 6 OF THE COUNCIL ON SCIENCE AND PUBLIC HEALTH (A-) Screening Mammography (Resolution 509-A-0, Resolve ) (Reference Committee E) EXECUTIVE SUMMARY Objectives. In November 009, the United States Preventive Services Task Force (USPSTF) updated its guidelines on routine screening for breast cancer. The updated recommendations are different from those of several other guideline-making groups and have contributed to the continuing debate about when routine screening mammography should begin and what its frequency should be. This report will highlight current screening mammography guidelines, explore the established benefits and harms of mammography, review the process by which the USPSTF developed its updated recommendations on screening mammography, and update the AMA s current policy recommendations. Data Sources. Literature searches were conducted in the PubMed database for English-language articles published between 000 and 0 using the search terms screening mammography, and mammography AND USPSTF, and mammography AND 40. To capture reports that may not have been indexed on PubMed, as well as news articles and press releases, periodic Google searches were conducted using the search terms mammography, mammography AND USPSTF, and mammography AND 40. Additional articles were identified by review of the literature citations in articles found in the PubMed and Google searches. Specific information on the USPSTF was obtained from its website. Results. Screening mammography reduces mortality from breast cancer, including in women younger than age 50 years. However, screening mammography carries harms such as falsepositive results that can lead to additional imaging and invasive biopsy procedures, and overdiagnosis that could lead to treatment in patients who may not benefit from it. The USPSTF considered the balance of benefits and harms using a commissioned targeted systematic evidence review of randomized clinical trials and a decision analysis that compared the expected health outcomes of starting and ending mammography at different ages and using annual and biennial screening strategies; it concluded (in part) that routine screening should begin at age 50 years and continue biennially until age 74 years. Several medical specialty societies, patient advocacy groups, and individuals offered either support for or opposition to the recommendations. Some groups have concurrently called for reform in the guideline development process. Conclusions. Mammography is a proven method for detecting breast tumors, with demonstrated reductions in mortality for women who undergo regular screening. Associated harms exist, which underlie differences in recommendations regarding the frequency and age at which to begin and end screening. Groups developing guidelines have placed different emphasis on these harms, resulting in varied conclusions about whether benefits outweigh harms, and whether that balance changes in different age groups. Mammography screening guidelines themselves regularly undergo review and update processes; the Council believes that it is appropriate for AMA policies referencing such guidelines to be reviewed and updated as well, and offers revisions to AMA policy H [Mammography Screening in Asymptomatic Women Forty Years and Older]. The foundation of the Council s recommendation is the notion that every woman age 40 years and older who wants a routine screening mammogram and whose physician believes it is clinically appropriate should receive one, regardless of her insurance coverage status. Action of the AMA House of Delegates 0 Annual Meeting: Council on Science and Public Health Report 6 Recommendations Adopted as Amended, and Remainder of Report filed.

2 REPORT OF THE COUNCIL ON SCIENCE AND PUBLIC HEALTH CSAPH Report 6-A- Subject: Presented by: Referred to: Screening Mammography (Resolution 509, A-0, Resolve ) Lee R. Morisy, MD, Chair Reference Committee E (Frederick R. Ridge, Jr., MD, Chair) INTRODUCTION Resolution 509-A-0, introduced by the Illinois Delegation, asked that our American Medical Association (AMA): () recommend that physicians and patients continue to follow the guidelines of the American Cancer Society regarding screening mammography and patient breast selfexamination; and () encourage government panels and task forces dealing with specific disease entities to have representation by physicians with expertise in those diseases. Resolve was referred for decision; Resolve was adopted. The Board of Trustees considered Resolve and referred it to the Council on Science and Public Health, asking for a report back on the issue of screening mammography, especially with regard to screening women ages years. Accordingly, this report will highlight current screening mammography guidelines, explore the established benefits and harms of mammography, review the process by which the United States Preventive Services Task Force (USPSTF) developed its updated recommendations on screening mammography, and update the AMA s current policy recommendations. METHODS Literature searches were conducted in the PubMed database for English-language articles published between 000 and 0 using the search terms screening mammography, and mammography AND USPSTF, and mammography AND 40. To capture reports that may not have been indexed on PubMed, as well as news articles and press releases, periodic Google searches were conducted using the search terms mammography, mammography AND USPSTF, and mammography AND 40. Additional articles were identified by review of the literature citations in articles found in the PubMed and Google searches. Specific information on the USPSTF was obtained from its website. BACKGROUND From , the USPSTF recommendations on breast cancer screening supported routine screening mammography, with or without a clinical breast exam, every - years for women age 40 years and older. These recommendations were similar to the recommendations of several other medical professional societies and cancer advocacy groups, including the American Cancer Society Action of the AMA House of Delegates 0 Annual Meeting: Council on Science and Public Health Report 6 Recommendations Adopted as Amended, and Remainder of Report filed.

3 CSAPH Rep. 6-A- -- page of (ACS), American College of Radiology (ACR), American Congress of Obstetricians and Gynecologists (ACOG), and the National Comprehensive Cancer Network (NCCN). In November 009, the USPSTF updated its guidelines on screening for breast cancer. These guidelines recommend against routine screening mammography in women aged years, and recommend biennial screening mammography in women aged years. The USPSTF concluded that the evidence was insufficient to recommend for or against routine screening mammography in women older than age 74 years. In December 009, the USPSTF updated the language of its recommendation regarding women under age 50 years to clarify its original and continued intent. That recommendation now states: The decision to start regular, biennial screening mammography before the age of 50 years should be an individual one and take patient context into account, including the patient's values regarding specific benefits and harms. The USPSTF also updated recommendations on clinical breast examination (CBE), self-breast examination (SBE), digital mammography, and magnetic resonance imaging (MRI), however this report will focus on the recommendations for screening mammography. RELEVANT AMA POLICY AMA policy strongly supports mammography screening for the early detection of breast cancer (see Appendix I). Policy H [Early Detection of Breast Cancer, AMA Policy Database] encourages recognition of mammography as an effective screening technique and additionally encourages education and awareness about breast self-examination. Policies H [Screening and Treatment for Breast and Cervical Cancer], H [Screening and Education Programs for Breast and Cervical Cancer Risk Reduction], and D [Mammography Screening for Breast Cancer] support funding for screening programs, including for low-income women; H additionally encourages educational programs to inform women about screening. With regard to recommendations directly addressing screening mammography in women between the ages of years, AMA policy is the following: H Mammography Screening in Asymptomatic Women Forty Years and Older. Our AMA strongly endorses the positions of the American College of Obstetrics and Gynecology, the American Cancer Society, and the American College of Radiology that all women have screening mammography as per current guidelines.. Our AMA favors participation in and support of the efforts of the professional, voluntary, and government organizations to educate physicians and the public regarding the value of screening mammography in reducing breast cancer mortality. 3. Our AMA advocates remaining alert to new epidemiological findings regarding age-specific breast cancer mortality reduction following mammography screening. 4. Based on recent summary data our AMA recommends annual screening mammograms and continuation of clinical breast examinations in asymptomatic women 40 years and older. 5. Our AMA encourages the periodic reconsideration of these recommendations as more epidemiological data become available. 6. Our AMA supports seeking common recommendations with other organizations. 7. Our AMA reiterates its longstanding position that all medical care decisions should occur only after thoughtful deliberation between patients and physicians. (CSA Rep. F, A-88; Reaffirmed: Res. 506, A-94; Amended: CSA Rep. 6, A-99; Appended: Res. 0, A-0) The original iteration of this policy was adopted in 988, based on the recommendations in Council on Scientific Affairs Report F-A The report recommended supporting annual screening mammography in women age 50 and older, and mammography screening every - years in

4 CSAPH Rep. 6-A- -- page 3 of women aged years. 3 The policy was updated in 999 by CSA Report 6-A-99, which recommended supporting annual screening mammography in asymptomatic women age 40 years and older. 4 In 00, with the adoption of Resolution 0-A-0, the policy was further amended to endorse the screening guidelines of ACOG, ACS, and ACR. CURRENT MAMMOGRAPHY SCREENING GUIDELINES Many organizations have developed or endorsed guidelines regarding screening mammography. The Table below summarizes the recommendations of several groups in this country, as well as those from the Canadian Task Force for Preventive Health Care 5 and Britain s National Health Service. 6 The USPSTF recommends routine screening mammography beginning at age 50 years and continuing biennially through age 74 years; the American Academy of Family Physicians (AAFP) endorses the recommendations of the USPSTF.,7 For women aged years, the USPSTF (with AAFP endorsing) and the American College of Physicians (ACP) recommend individual patient assessment for breast cancer risk, along with patient education about the benefits and limitations of mammography, as the basis for a decision to screen.,7,8 ACOG, ACR, ACS, and NCCN recommend annual routine screening mammography beginning at age 40 years. 9- ACOG, ACS, and NCCN include in their guidelines a recommendation to discuss with women the predictive value of mammography and its limitations. 9,, ACOG states that based on individual risk, biennial screening may be appropriate for some women. 9 ACOG, ACR, ACS, and NCCN guidelines do not specify an age at which screening should end. While NCCN states that the appropriate upper age limit has not yet been determined, ACR recommends continuation until life expectancy reaches less than five to seven years, 0 and ACS recommends continuation as long as the patient is in good health. ACOG notes that women 75 years or older should, in consultation with their physicians, decide whether or not to continue mammographic screening. 9 Organization (year recommendation updated) Age at which routine screening should begin Frequency Age at which routine screening should end AAFP (009) a 50 Biennial 75 ACOG (0) 40 (with discussion c ) Annual (Biennial may be Not specified appropriate for some) ACR/SBI b (00) 40 Annual Life expectancy <5 7 years ACS (003) 40 (with discussion c ) Annual As long as patient is in good health NCCN (0) 40 (with discussion c ) Annual Not yet established USPSTF (009) 50 Biennial 75 CTFPHC (0) 50 Triennial 75 NHS (0) 50 (expanding to 47) Triennial 70 (expanding to 73) Table: Screening mammography recommendations of several groups. Abbreviations are as follows: AAFP: American Academy of Family Physicians; ACOG: American Congress of Obstetricians and Gynecologists; ACR: American College of Radiology; SBI: Society of Breast Imaging; ACS: American Cancer Society; NCCN: National Comprehensive Cancer Network; USPSTF: United States Preventive Services Task Force; CTFPHC: Canadian Task Force for Preventive Health Care; NHS: National Health Service (Britain) a. The AAFP endorses the USPSTF s recommendations b. ACR and SBI have joint recommendations. c. Recommendation includes the discussion of the predictive value and limitations of mammography.

5 CSAPH Rep. 6-A- -- page 4 of A survey of the International Breast Cancer Screening Network shows that 5 of 9 member countries recommend screening beginning at age 40 years, with most screening biennially. 3 The recommendations of the different countries are, by and large, based on the same data, but reflect a difference of opinion in data interpretation. 3 It is important to note that the guidelines discussed in this report are for routine screening mammography, i.e., mammography for women who are at average risk for breast cancer. They are not appropriate for women at increased risk due to underlying genetic mutations (such as BRCA or BRCA), family history, previous chest radiation, or other risk factors; guidelines for women at increased risk are substantially different., BENEFITS AND HARMS OF SCREENING MAMMOGRAPHY Breast cancer is the most common cancer in women in the U.S., with more than 00,000 women receiving a diagnosis of invasive breast cancer each year and nearly 40,000 dying. 4 The average woman s lifetime risk of developing breast cancer is in 8, or %, 4 however factors such as age, family or personal history of cancer, dense breasts, and previous exposure to chest radiography can increase risks. 5 In the U.S., digital mammography has rapidly replaced the older method of film mammography. 6 Though mammography is the most reliable breast cancer screening tool for the general population, it carries potential harm along with its benefits. Recommendations regarding screening frequency and age of initiation are based on the balance of benefits and harms. Benefits of screening mammography Mortality reduction. There is wide agreement that screening mammography leads to a reduction in breast cancer mortality, 7 although disagreements exist about how to calculate such reductions. Randomized controlled trials (RCTs) have estimated the reduction in mortality across all age groups to be approximately 5-30%, 8- while observational and modeling studies have estimated mortality reduction across all age groups to be higher, with a range of 30% to more than 40%. 3-6 In RCTs, mortality reduction is based on the number of women invited to screen, rather than those who have actually undergone screening in the trial. This number invited to screen includes those women who are part of the screening arm of the trial but who decline screening. Those who fit into this category and who also die of breast cancer will be counted in the larger number of women in the screening arm that died of breast cancer. 6 Based on this method, noncompliance to the screening protocol potentially underestimates the mortality reduction derived from screening. 6 Similarly, women who are assigned to the control, non-screening arm sometimes seek mammography on their own, skewing the potential mortality reduction downward. 6 There have been few RCTs designed to determine mortality reduction from mammography screening in specific age groups; estimates have been derived from subanalyses of trials designed for other outcomes. Pooled data from RCT subanalyses show mortality reduction from mammography screening to be greatest in women aged years (approximately 3%). 8 For women aged years and years, pooled data show mortality reduction to be 5% and 4%, respectively. 8,,7-3 Although these values appear to indicate a similar mortality reduction for both of these age groups, it should be noted that estimated reductions are based on relative risk (risk of breast cancer mortality in women of a particular age group who undergo mammography versus those in the same age group who do not undergo mammography). Because a woman s risk for breast cancer increases sharply with age, absolute mortality risk reduction (reduction in the overall risk of breast cancer mortality) from screening is greater for women aged years than that for women aged years.,8 Mortality reduction estimates for women age 70 years and older are lacking because of insufficient data. 8

6 CSAPH Rep. 6-A- -- page 5 of Subanalyses of trials designed to estimate benefit across larger age groups, as well as more recent retrospective studies, have shown benefits for women aged years who undergo screening mammography.,6,30 Between years of age, tumors detected by mammography are smaller with less nodal metastasis (compared to those tumors detected without mammography), and 5-year and disease-free survival are improved. 33 Additionally, a 00 study showed that mammography in women younger than age 50 years with a family history of breast cancer increases cancer detection, reduces risk of advanced stage disease, and is associated with lower mortality and higher 0-year survival from invasive cancer. 34 Based on analyses of breast cancer mortality reduction before and after the implementation of screening programs, some argue that the observed reduction is only partially due to screening, with the rest due to improved therapy and management of breast cancer disease and to changes in staging techniques. 5,35,36 However, this is refuted by others. In regions without formal screening but with access to improved treatments, the mortality rate did not decrease until screening was introduced. 37,38 It is possible that the mortality reduction associated with screening mammography could be greater. Only approximately 65% of women age 40 years or older report having undergone screening mammography within the last two years. 39 Increasing adherence to recommendations could potentially increase the number of women in whom cancer is detected early, leading to greater mortality reduction.,39 Harms of screening mammography Although there is broad agreement that screening mammography reduces mortality from breast cancer, it is not a perfect tool. Along with the intended early detection of invasive breast cancer, mammography carries with it potential harms, such as false-positive results, overdiagnosis, and exposure to radiation. False-positive results. A false positive is defined as an abnormal screening mammography result that does not end in a diagnosis of invasive carcinoma or ductal carcinoma in situ (DCIS) within one year of the screening examination. 40 The reported specificity of mammography is 94-97%. 0,4 In other words, 94-97% of mammograms correctly rule out the presence of disease in disease-free individuals. Though this specificity appears to be high, it must be considered in the context of the number of mammograms performed. More than 33 million screening mammograms are performed in the U.S. each year. 4 Taking into account the annual incidence of breast cancer (approximately 4 cases per 00,000 women), 43 the reported specificity implies that every year, approximately - million women receive an abnormal mammography result that will turn out not to be breast cancer. Many of these women will undergo further imaging and invasive procedures. 44 A 0 study, designed to address limitations in previous estimates of false-positive rates, found that after 0 years of annual screening, the probability of receiving a recall (recommendation for immediate follow-up imaging) is 6.3%; this probability drops to 4.6% for 0 years of biennial screening. 44 These estimates are similar whether screening begins at age 40 or 50 years. Older studies report that false-positive mammograms occur in -49% of all women after 0 mammography examinations, and in up to 56% for women aged years. 8 The probability of a false-positive biopsy recommendation (recommendation for biopsy, fine-needle aspiration, or surgical consult after imaging work-up) is 7-9% after 0 years of annual screening and 4-6% after 0 years of biennial screening. 44 While biennial screening appears to decrease the probability of a

7 CSAPH Rep. 6-A- -- page 6 of false-positive mammography result, it may be associated with an increase in the probability of a late-stage cancer diagnosis. 44 Many women who have been recalled for further screening become distressed, and some report persistent anxiety despite eventual negative results. 8,49 Others report only transient anxiety. 8,37 False-positive results appear to affect breast cancer-specific distress, anxiety, apprehension, and perceived risk rather than general depression and anxiety. 8,50 False-positive results can also affect adherence to screening recommendations. In a 0 study, women who received a false-positive result were less likely to return for routine screening compared with women who received negative results. 5 However, reattendance improved with the number of completed screening participations, suggesting that abnormal results in younger women (who have completed relatively few screens) are more likely to negatively impact reattendance than in women who have undergone several routine screens. 5 Variation in screening mammography specificity has been noted among physicians and facilities. For example, recall rates are lower and specificity rates higher among radiologists who have more years of experience interpreting mammograms. 5,53 Higher specificity is seen at facilities that offer screening mammography alone (versus those that offer both screening and diagnostic mammography), have a breast-imaging specialist interpreting mammograms, and conduct audit reviews two or more times each year. 54 AMA policy (H Safety and Performance Standards for Mammography; see Appendix I) supports high quality standards of performance for those administering and interpreting mammograms, including evidence of appropriate training and competence for professionals. Overdiagnosis. Overdiagnosis is the detection of cancer that would not have clinically surfaced in a person s lifetime, usually because of lack of progressive potential. 4 Overdiagnosis is easily confused with false-positive results, i.e., a positive screening result that is subsequently determined not to be cancer. In contrast, an overdiagnosis represents a case in which the pathological criteria for cancer has been fulfilled. 55 Stable disease including some DCIS, indolent cancers, and slowgrowing tumors are thought to be most commonly overdiagnosed by mammography. 55,56 Some reports have concluded that a small percentage of mammography-detected cancers may spontaneously regress, although others have criticized this assertion Evidence for overdiagnosis comes from RCTs designed to demonstrate the benefit of mammography. In these trials, women are randomly assigned to screening mammography and non-screening mammography arms; since the assignments are random, the number of breast cancers that develop over time should be the same in each group. 59 In the group receiving screening mammography, the number of women receiving breast cancer diagnoses will initially be higher than in the non-screening group, since the mammograms will detect tumors too small to be detected otherwise. With time, as the small tumors in women in the non-screening group grow and become detectable, the number of breast cancer diagnoses should become similar to those in the screening group. However, some trials have shown that breast cancer diagnoses in the screening group are persistently higher, even after many years. This persistent difference represents overdiagnosis. 59 Quantification of overdiagnosis is difficult; it is not ethically possible to set up prospective clinical trials to determine which cancers will remain indolent if left untreated. 60 Therefore, the proportion of mammography-detected breast cancers that are estimated to be overdiagnoses is widely variable, ranging between -30%; estimates are derived from screening programs in several countries that are statistically difficult to combine. 8 Observational and modeling studies have attempted to

8 CSAPH Rep. 6-A- -- page 7 of narrow the range. For example, a 0 study used data from different geographic regions in Norway, where screening mammography began at staggered times over a nine-year period. 6 By comparing breast cancer incidence in regions with a screening program to incidence in regions that had yet not implemented screening, the study estimated that 5-5% of mammography-detected breast cancers were overdiagnoses. 6 Within different age groups, modeling studies have shown only small differences in the rate of overdiagnosis. In general, the risk for overdiagnosis increases with age, likely because in older age groups, rates of competing causes of mortality increase. 4 The difficulty in accurately estimating rates of overdiagnosis has led to arguments that the estimates are artificially high, and are complicated by follow-up times, lead-time, and changes in breast cancer incidence over several years. 6 Overdiagnosis is regarded by some as the most serious harm associated with mammography; 59 at the time of diagnosis, clinicians cannot know who has been overdiagnosed, so all are treated for potentially lethal cancer. 55,56 These patients will not benefit from treatment and almost certainly will be harmed. 55 A perceived benefit of mammography screening is that it reduces the need for mastectomies and increases the potential for breast-conserving treatment. 63 However, a 3% increase in breast surgery and 0% increase in mastectomy for women exposed to screening has been reported. 9 A 0 Norwegian study corroborated these findings, and concluded that overdiagnosis is likely to have contributed to the increases in surgical intervention. 63,64 Other studies have reported no increase in the rate of mastectomy. 65,66 Radiation exposure. Little evidence exists to suggest that low-dose radiation exposure from mammography is a significant risk. 8 Widely-ranging cumulative radiation doses of Gy are thought to significantly increase the risk for breast cancer; 67 the average dose for a bilateral, two-view mammogram is 7 mgy or less, 68,69 and for women aged years, annual mammography screening for 0 years (with potential additional imaging) exposes the individual to approximately 60 mgy. 67 The number of radiation-induced breast cancer deaths associated with biennial screening between the ages of years has been modeled at.6 per 00,000 women screened. This model also predicts that extending the biennial screening period to women between the ages of years results in 3.7 radiation-induced breast cancer deaths per 00,000 women. 69 These rates are considered negligible, with screening benefits far outweighing the risk of radiation exposure. 8,69 For comparison, the ratio of breast cancer deaths prevented by mammography to the number of deaths induced by radiation exposure is 684: for women aged years, and 349: for women aged years. 69 Special consideration of the effects of radiation exposure should be given to women who have previously undergone diagnostic chest radiographs or had therapeutic radiation for other cancers. These women are at increased risk for cancer since cumulative radiation exposure is increased. 70 THE USPSTF AND ITS RECOMMENDATIONS FOR SCREENING MAMMOGRAPHY Background The mission of the USPSTF is to review the scientific evidence for clinical preventive services and to develop evidence-based recommendations for primary care physicians as well as the broader health care community. 7 Congress codified the USPSTF as an independent body in 998. Though the Agency for Healthcare Research and Quality (AHRQ) is mandated to convene the USPSTF, its sole role is to support the USPSTF by providing meeting space, organizing conference calls,

9 CSAPH Rep. 6-A- -- page 8 of managing contracts for systematic reviews, and providing staffing. 7 No individual at AHRQ has a vote in the recommendations, or otherwise influences the priorities or decisions of the USPSTF. 7 The USPSTF comprises 6 members who serve terms of 4-6 years; members are appointed by the AHRQ director based on recommendations developed by the USPSTF Chair and Vice-Chair following a public nomination process. 7 Members are experts in primary care and preventive health-related disciplines, and collectively possess expertise in evidence-based clinical research, screening, clinical epidemiology, behavioral science, health services research, outcomes and effectiveness in clinical preventive medicine, and decision modeling. 7 The USPSTF does not deliberately seek out task force members who are experts on specific topics; experts bring substantial knowledge regarding guideline development processes but also may retain inherent biases. 7,73 It is sometimes difficult for experts to fairly assess and critique studies that they or their colleagues have conducted, contradict beliefs entrenched since training, and recommend against services that may benefit themselves or their specialties. 7 Also, many experts in specific topic areas lack training in epidemiology and biostatistics. 7 The USPSTF is considered unique in that it convenes primary care providers and scientists with skills in objectively critiquing studies without preconceived views or a stake in the outcome. 7 The USPSTF follows a detailed protocol for guideline development. 74 For each topic under consideration, an AHRQ evidence-based practice center conducts a systematic review of the evidence, which enables a subcommittee of the USPSTF to develop estimates of the magnitude and certainty of benefits and harms. These estimates are extensively reviewed by the full USPSTF in order to reach consensus and vote on recommendations. Cost and cost-effectiveness are not considered in the guideline development process. 7 A full explanation of the USPSTF s evidence grading and subsequent recommendation system is published on the USPSTF website. 74 Subspecialist experts in the disease at hand, as well as partner organizations, are asked to review and comment on USPSTF work at three points in the recommendation development process:. the initial analytic framework and key questions that drive the systematic review;. the systematic review itself; and 3. the draft recommendation statement. USPSTF partner organizations that are also members of the AMA Federation of Medicine are AAFP, ACOG, ACP, the American College of Preventive Medicine (ACPM), the American Academy of Pediatrics, and the American Osteopathic Association. Recommendations for screening mammography Plans for the update of the 00 USPSTF recommendations on screening mammography began in late 006. In 007, the USPSTF commissioned two reviews: a targeted systematic evidence review of the benefits and harms of screening 75 and a decision analysis based on modeling techniques that compared the expected heath outcomes of starting and ending mammography at different ages and using annual and biennial screening strategies. 4 The systematic review excluded studies other than RCTs and systematic reviews or those without breast cancer mortality as an outcome. 8,75 The systematic review included analyses of evidence regarding CBE, SBE, digital mammography, and MRI, but this section will focus on the evidence analyzed to develop recommendations on screening mammography. In its 009 update, the USPSTF recommended against routine screening mammography for women aged years, and instead recommended an individualized decision to screen during this time period. This recommendation is partially based on findings in the commissioned systematic review. 8 The systematic review was carried out by the Oregon Evidence-based Practice Center, funded by AHRQ. Prior to its finalization, the draft report was reviewed by 5 experts not

10 CSAPH Rep. 6-A- -- page 9 of affiliated with the USPSTF. These reviewers included one oncologist, an expert in modeling, two radiologists, one breast surgeon, and three physician/epidemiologists. 76 The names of the reviewers are included in the full systematic review available on the National Library of Medicine website. 75 Mortality reduction was considered an important outcome in the formation of the recommendations. The systematic review estimated the mortality reduction for women aged years, years, and years to be 5%, 4%, and 3% respectively. 8 These estimates are similar to those established in the USPSTF s 00 systematic review, but include new data from an update of a previously completed trial, 30 and another clinical trial completed after 00.,0,3 Since these mortality reduction estimates are based on relative risk, the USPSTF considered calculations of the number needed to invite for screening to prevent one death from breast cancer, which more clearly explains mortality reduction. The number needed to screen calculation is based on absolute risk, so it takes into account the background risk for breast cancer. 77,78 This number can more clearly reflect the benefit of mammography in each age group since it includes the increasing absolute risk of breast cancer with advancing age. The number needed to invite for screening (to prevent one death) is 904 for women aged years, 339 for women aged years, and 377 for women aged years. In addition to the mortality reduction benefit associated with mammography, the USPSTF considered harms. In some studies, the probability of receiving a false-positive mammography result is slightly higher in women aged years. 8 A false-positive mammography result often leads to additional imaging, and after several years participating in a screening program, nearly 0% of women receive a false-positive biopsy recommendation. 44 Though the range of reported overdiagnosis is large, between -30%, and therefore difficult to estimate precisely, it is a risk that many agree is serious, since it leads to treatment that may not be necessary. 8 Radiation exposure was not considered to be a serious risk of screening mammography, except for the small percentage of the population previously exposed to chest radiography and therapeutic radiation. 70 The USPSTF-commissioned decision analysis compared the expected health outcomes of starting and ending mammography at different ages and using annual and biennial screening strategies. 4 For the screening models compared, biennial screening retains 70-99% of the reduction in mortality that occurs with annual screening, depending on the age range for screening. 4 The models predict that beginning screening at age 40 years yields an additional 3% mortality benefit compared with beginning screening at age 50 years. 4 This additional mortality benefit is the same with either annual or biennial screening beginning at age 40 years. 4 Extending screening to age 79 years yields an additional 8% or 7% mortality benefit compared with screening programs ending at age 69 years, for annual and biennial screening, respectively. 4 If the two strategies are compared, these data indicate that greater mortality reduction could be achieved by continuing screening past age 69 years rather than by initiating it at age 40 years. However, if life-years gained is considered, models show that initiating screening in younger women rather than extending screening in older women results in more benefit; this is not surprising since younger women have longer life expectancies than older women. Annual screening between the 9 year period comprising ages years yields a median of 33 life-years gained per 000 women screened, whereas annual screening between the 9 year period comprising ages years yields a median of 4 life-years gained per 000 women screened. 4 Biennial screening with these parameters yielded 5 and 3.5 life-years gained in the two groups, respectively. The decision analysis also compared the harms associated with different screening models. Annual screening between ages years yields,50 false positive results for every 000 women screened over the 9 year period, almost twice as many as that of a biennial screening period. 4

11 CSAPH Rep. 6-A- -- page 0 of Consequently, many more women who are screened annually will undergo biopsy compared with those who are screened biennially. 4 The models also predict an increase in the risk of overdiagnosis as age increases. Overall, initiating screening at age 40 years (compared to age 50 years) had a smaller effect on overdiagnosis than extending screening beyond age 69 years. 4 Overdiagnosis risk was smaller with biennial screening, but by less than half. 4 The USPSTF studied the balance of benefits and harms of mammography, as well as the results of the systematic review and the decision analysis study, to develop its final recommendations. It concluded that compared with initiating screening at age 50 years, screening mammography provides a small benefit when performed annually in women aged years, but is more likely to be accompanied by false-positives and overdiagnosis, resulting in a smaller net benefit.,7 The ages at which the balance of benefits and harms becomes acceptable to individuals and society are not clearly resolved by available evidence. 9 Because of the small net benefit, the USPSTF concluded that mammography in women aged years should not be automatic, but should instead be initiated as a result of an individual decision based on the woman s specific clinical situation, preferences, and values regarding the potential benefits and harms.,7 REACTION TO USPSTF RECOMMENDATIONS The 009 USPSTF screening mammography recommendations were met with opposition by several medical specialty societies, public advocacy groups and individuals in the medical community. ACR stated that the USPSTF recommendations were ill-advised and would result in countless unnecessary breast cancer deaths each year. 79,80 ACOG, ACS, the Radiological Society of North America, the Society of Breast Imaging, the American Society of Breast Disease, and other groups also publicly stated opposition to the recommendations Most reiterated support of guidelines that recommend routine screening mammography beginning at age 40 years. Several publications addressing perceived flaws in the interpretation of data by the USPSTF have appeared in peer-reviewed journals. 3,6,38,85-88 Among the criticisms of the USPSTF process was reliance on only RCTs in the evidence review, with the exclusion of additional observational studies showing higher mortality benefit and reduced numbers needed to screen. 79 Several studies, including some RCTs, did not meet the USPSTF s strict inclusion criteria; others received only a grade of fair for their shortcomings. 74,75 Another criticism was the use of the number needed to invite for screening value, rather than the number actually screened. 6 The USPSTF reported that the level of participation in the trials was high, and that data from trials with lower participation rates was graded as lower quality. 7,74,75 The USPSTF also reported that the use of only participating women, rather than those who were merely invited to screen, yielded only a slightly higher benefit. 7 In contrast to the opposition, several organizations, including those representing primary care physicians and public health providers, expressed public support for the 009 USPSTF recommendations. In a letter to members of Congress, health care organizations, including the AAFP, ACP, and ACPM defended the recommendations. 89 The AAFP also joined with four of its affiliate groups to urge the Secretary of the Department of Health and Human Services to reject calls to remove the USPSTF recommendations from the AHRQ website. 90 Advocacy groups, including the National Breast Cancer Coalition, Breast Cancer Action, and the National Women s Health Network also publicly supported the USPSTF recommendations Media coverage of the USPSTF recommendations was often controversy-oriented A recent study reported that more than half of media reports about the recommendations took an unsupportive stance; nearly 70% of reports included the belief that delayed screening leads to

12 CSAPH Rep. 6-A- -- page of more breast cancer and related deaths or concern over cost and government rationing of health care. 98 Seventeen percent of the reports took a supportive stance, based on beliefs that the recommendations were based on science and that there is potential harm in mammography. 98 Not surprisingly, laywomen who had, or currently have, breast cancer were angered by the recommendations, strongly believing that mammography saved their lives. 99 The opinions of women who have not experienced breast cancer also were strongly influenced by media coverage, with women who had viewed commentary that was critical of the USPSTF guidelines more likely to overestimate individual risk for breast cancer and feel uncomfortable about delaying mammography until age 50 years, compared to those who viewed commentary that supported the USPSTF guidelines. 00 At the time that the recommendations were released, the country was deeply involved in the debate about health care reform. Since the USPSTF is convened by a government agency (AHRQ), several media outlets and others expressed serious concern that the recommendations would be binding in government health care policy. Several journal publications expressed the opinion that USPSTF is an opponent of screening and that its recommendations were intended to restrict patient access to mammography. 6,38,86 Others joined in suggesting that the recommendations would directly affect costs and insurance coverage for breast cancer screening, and calls were made for Congress to intervene. In response, in early December 009, the Senate passed amendments to its proposed health care reform legislation: one requiring the federal government to effectively ignore the new recommendations, and the other guaranteeing no-cost breast screening for women in their 40s. These provisions were signed into law in 00 as part of the Affordable Care Act. INDIVIDUAL AND RISK-BASED SCREENING The USPSTF is not the first group recommending an individualized, risk-based approach to mammography screening in women aged years, 8 but the attention paid to the mammography recommendation has highlighted consideration of that approach. Individualized screening refers to screening mammography at an age and frequency decided upon by both physician and patient, based on the physician s assessment of patient clinical factors that influence breast cancer risk and the patient s values regarding the balance of benefits and harms of screening mammography. Data suggest that women themselves want to be involved in the decision to initiate screening mammography, and often request specific information prior to their first mammogram, including information about benefits and harms. 0 Women acknowledge anxiety about false positives, but show little awareness of overdiagnosis. 0 Physicians have an ethical obligation to educate women with balanced information appropriate to the desire expressed by each patient for such information. 0 Model physician-patient dialogue and patient decision aids have been developed as resources to support the shared decision-making underlying the individualized screening approach Some argue that the individualized risk-based screening approach will fall short in effectively detecting early cancer. A large percentage of cancers are diagnosed in women with no apparent risk factors, suggesting that relying on the identification of personal or family risk factors to indicate the need for mammography will miss many cancers that could have been detected by mammography. 06,07 Also, randomized data are lacking to support a risk-based approach between the ages of years since no RCTs have stratified participants by risk. 06 However, there are hints that a risk based approach may be effective. In a recent single arm (non-controlled) study, women ages years at intermediate risk for breast cancer (those with at least one first-degree relative with breast cancer) who were screened annually had smaller tumors that were less likely to be node-positive when compared to control groups from other studies. 34 Additionally, a meta-

13 CSAPH Rep. 6-A- -- page of analysis and systematic review examining several risk factors found that breast cancer in a firstdegree relative and extremely dense breasts were associated with increased risk in women ages years. 08 An accompanying modeling study found that for women with either one of those two risk factors, biennial screening mammography beginning at the age of 40 years has the same balance of benefits and harms as that for biennial screening mammography beginning at age 50 years in women without those risk factors. 09 The individualized approach relies heavily on the identification of red flags in a patient s family history, yet many patients do not receive adequate familial cancer risk assessment in the primary care setting. 0-3 Further, a patient s family history will change over time as family members health status changes. Clinically relevant family history changes substantially during early and middle adulthood (between the ages of years), particularly for breast cancer. 0 If a patient s family history is not updated adequately during those years, risk factors that would indicate a need for more intensive screening will be missed. 0 Some physicians also do not follow recommendations for referral of women for high-risk cancer genetic counseling, suggesting that estimation of breast cancer risk by these physicians is faulty. 4 This behavior may reflect a misunderstanding of what constitutes high risk, since definitions are variable.,8,5,6 GUIDELINE REFORM The controversy stemming from the 009 USPSTF recommendations has brought attention to the process of guideline development. ARHQ s National Guideline Clearinghouse contains close to,700 clinical practice guidelines, and the number of groups issuing guidelines has proliferated, along with substantially different development methodologies. 7 The Clearinghouse was originally created by AHRQ in partnership with the AMA and the American Association of Health Plans (now America s Health Insurance Plans). With the growth in the number of guidelines being developed, physicians, consumer groups, and other stakeholders have expressed concern about the quality of the processes used to develop guidelines, and the resulting questionable validity of many guidelines. 7,8 Concerns stem from limitations in the scientific evidence base, a lack of transparency in the methodologies used by guideline-developing groups, conflict of interest among guideline-developing group members and funders, and uncertainty regarding how to reconcile conflicting guidelines. 7 Additionally, significant variability in the recommendations of guidelines can lead to confusion and frustration on the part of health care providers and patients. 9 Specific to mammography guidelines, a recent study suggests that guideline development reform is needed. The study assessed the quality of guidelines that provide recommendations on mammography screening in asymptomatic women aged years, and concluded that both the evidence reviews underlying the guidelines, as well as the guidelines themselves, were of vastly different quality. 9 Based on quality assessment instruments, the study assigned an overall assessment for use in clinical practice to each of the guidelines. Of the guidelines studied, only three received strongly recommend or recommend assessments. 9 The remaining guidelines were found to have deficiencies in their development processes, and were given unsure or would not recommend assessments. 9 In response to concerns that the guideline development process is widely variable, thus leading to guidelines that are variable in quality, the Institute of Medicine (IOM) recently undertook a project to define standards for guideline development. 7 The standards, released in Spring 0, promote the development of unbiased, valid, and trustworthy guidelines that incorporate a grading system for characterizing the quality of evidence and strength of clinical recommendations. 9 Standards are focused on establishing transparency, managing conflicts of interest, composition of the development group, systematic review use, evidence strength, articulation of recommendations,